The issue of measurement uncertainty estimation arising from sampling for model and empirical approaches is raised. The instances, when uncertainty estimation arising from sampling is reducing to estimation of a random component, are given. Usage of a classic ANOVA in general is validated. The methodology for obtaining more accurate reliability of estimations and increasing the number of repeated measurements is introduced. The example of given methodology is given.
sampling, uncertainty, ANOVA, sample, power of test
"Osoblyvosti otsiniuvannia nevyznachenosti vymiriuvannia, poviazanoi z vidborom prob" ,
Information Processing Systems,